Professor Bernard Veldkamp
University of Twente

Professor Bernard Veldkamp is Head of the Department of Research Methodology, Measurement, and Data-Analysis, Faculty of Behavioral, Social and Management Sciences, University of Twente. He specializes in research methodology and data analytics. His interests focus on optimization, text mining, and computer-based assessment. His work spans a range of issues in educational, psychological, and health sciences, from the development of new methods/models for the design and construction of (adaptive) psychological and educational tests, to the development of data mining models for analyzing verbal data and large datasets in fraud detection. He founds his research in Psychometrics, Operations Research, Data Mining, and Statistics.

In the old days, carefully constructed exams were used to collect the data that were needed to estimate the ability level of the students and to make pass/fail decisions. Efforts were made to minimize noise in the data and measurement theories like item response theory were developed and optimized. Nowadays, we are flooded by data about their performance coming from social media, sensors, log files, the internet and many other sources. The question arises how to master this data mass and how to use it for educational measurement. In this presentation, some examples are provided about how these new data sources can be used to answer existing and new research questions, a blueprint for the use of data analytics will be presented, it will be applied to the problem of test development. Finally, the attributed value of social data analytics will be discussed.